{"id":"W4396832846","doi":"10.1080/01639374.2024.2347210","title":"Duplicate Records in WorldCat for 20th-Century American, British, and Canadian Books: A Comparison of Duplication Rates and Causes","year":2024,"lang":"en","type":"article","venue":"Cataloging & Classification Quarterly","topic":"Digital Humanities and Scholarship","field":"Arts and Humanities","cited_by":1,"is_retracted":false,"has_abstract":true,"ca_institutions":"Concordia University","funders":"","keywords":"Library science; Cataloging; History; Computer science; Information retrieval; Genealogy; World Wide Web","routes":{"ca_aff":true,"ca_fund":false,"ca_venue":false,"about_ca":true,"invisible_to_affiliation_only":false},"retraction":null,"screen":null,"direct_labels":[],"prediction":{"model_version":"codex-gemma-dda1882f352a","candidate_categories":["scholarly_communication"],"consensus_categories":[],"category_scores_codex":[0.0002189524,0.000138206,0.0002379504,0.0002650702,0.000201919,0.0010541,0.000113532,0.00003676007,0.00003184172],"category_scores_gemma":[0.00002913844,0.0001521118,0.00003878906,0.00008211062,0.0004439459,0.0005443638,0.000008582777,0.0001270632,0.000008263088],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.00009321987,"about_ca_system_score_gemma":0.00008728458,"about_ca_topic_candidate":true,"about_ca_topic_consensus":true,"about_ca_topic_score_codex":0.04007502,"about_ca_topic_score_gemma":0.4250276,"domain_scores_codex":[0.998873,0.00002543758,0.0003884543,0.0003550932,0.00008331358,0.0002747196],"domain_scores_gemma":[0.9993204,0.0001657238,0.0001161479,0.0001760086,0.00009832052,0.0001233794],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"design_other","study_design_gemma":"not_applicable","study_design_scores_codex":[0.00003927591,0.0001097797,0.01172534,0.0005374683,0.00005901478,0.000007164316,0.05810201,8.696979e-7,0.0004413363,0.4155936,0.009323189,0.5040609],"study_design_scores_gemma":[0.0004680235,0.000621078,0.09061395,0.0005786293,0.00007070995,0.00002133919,0.05604687,0.003503249,0.00005823495,0.006755579,0.8406307,0.0006316426],"study_design_candidate":"not_applicable","study_design_consensus":null,"genre_codex":"empirical","genre_gemma":"empirical","genre_scores_codex":[0.985783,0.005250168,0.0000772883,0.001515521,0.0002932304,0.0006019028,0.0004349238,0.0001080724,0.005935886],"genre_scores_gemma":[0.9974526,0.0001216305,0.00004578847,0.0001199176,0.0001054791,0.0001646248,0.000347709,0.00002391333,0.001618316],"genre_candidate":"empirical","genre_consensus":"empirical","teacher_disagreement_score":0.8313075,"threshold_uncertainty_score":0.9999829,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.06085712074718828,"score_gpt":0.2895339395762924,"score_spread":0.2286768188291042,"validation_status":"score_only:v0-immature-baseline","note":"Baseline scores from an immature model (maturity gate not passed). Scores rank; they never assert a category."}}